This is totally a thing people do, but it can be messy because 80% of the answer to "what's driving votes" is always "partisanship"
14.10.2025 22:17 β π 2 π 0 π¬ 0 π 0@charlotteeffect.bsky.social
Data witch for good causes and bad datasets
This is totally a thing people do, but it can be messy because 80% of the answer to "what's driving votes" is always "partisanship"
14.10.2025 22:17 β π 2 π 0 π¬ 0 π 0Nah I think that read is about right, just complicated because IRL you are dealing with both what voters like AND what will actually get covered
14.10.2025 22:16 β π 0 π 0 π¬ 0 π 0Very important to periodically step back from the niche factional fights and remember that you mostly agree and everyone is doing their best. In person meetings also very good for this!
14.10.2025 19:08 β π 1 π 0 π¬ 1 π 0Every time I get frustrated by chaos in the broad liberal coalition, Republicans go and do some shit like this and remind me that everyone i argue with is at least trying to make the world better, usually pretty nice, and not nearly this gross
14.10.2025 19:01 β π 7 π 1 π¬ 2 π 0Sorry for all the text, lol, I have survey thoughts and feelings
14.10.2025 18:29 β π 0 π 0 π¬ 1 π 0Imo the revealed preference is usually that you get more juice out of an issue that is surprising to be associated with a given partisanship, but it's super sensitive to the framing. Using messages ("a candidate who says blah") is better but even more sensitive
14.10.2025 18:28 β π 2 π 0 π¬ 1 π 0I'm currently doing zero modeled out results for Searchlight for transparency reasons , but it's common for folks to run this as a max diff or just toss it into a regression. Works okay!
14.10.2025 18:26 β π 0 π 0 π¬ 1 π 0Hm, like a "D who is focused on [issue]/R who is focused on [other issue]" thing? I've tried it in other work, it's interesting but I don't love it. Mostly because voters can interpret issues very differently when they have a partisanship attached
14.10.2025 18:26 β π 2 π 0 π¬ 1 π 0Oof. Yikes! Not good!
09.10.2025 20:42 β π 0 π 0 π¬ 0 π 0The #1 piece of advice that TransLink's CEO offers other North American transit agencies:
Provide frequent service
This goes beyond policy choices into vibes- what sort of life do Americans think each party is promoting? Mostly, not a very good one
Full writeup here from me @searchlightinst.bsky.social www.searchlightinstitute.org/research/che...
There's skepticism that *either* party wants you to have the markers of a good life, but it's worse for Democrats (even among Democratic voters)
09.10.2025 14:41 β π 1 π 1 π¬ 1 π 0People don't believe that Democrats want them to have a good life
09.10.2025 14:41 β π 0 π 1 π¬ 2 π 1Today: arguing that Sherrill might lose in NJ, and also why you probably shouldn't put too much faith in that guess (or any guess). Come stare at graphs, on the blog
open.substack.com/pub/cauldron...
This whole NY policy debate kind of blows my mind given that I recently disconnected and reconnected our gas range. It was a little annoying (like, a good wrench, some leverage, the right pipe tape) but not *hard*.
01.10.2025 16:13 β π 3 π 0 π¬ 0 π 0Great poll result from a very good @lakshya.splitticket.org piece today in The Argument. A lot of people think AI has been good for them but bad for others. www.theargumentmag.com/p/chatgpt-an...
30.09.2025 11:18 β π 17 π 6 π¬ 1 π 3Many such cases
26.09.2025 03:08 β π 1 π 0 π¬ 0 π 0Analytics and polling folks: I'll be at the Harvard political analytics conference this Friday if you want to get coffee or something
23.09.2025 19:09 β π 2 π 0 π¬ 0 π 0Oh totally. Voters love a villain. There's probably something clever you can do with framing new housing as breaking investor control but like oof it's messy!
23.09.2025 18:37 β π 3 π 0 π¬ 1 π 0I will commission the Searchlight graphic designers for a giant sign that says "do it you coward", it'll be multi purpose
23.09.2025 18:30 β π 19 π 2 π¬ 2 π 0I mean, Sirota is being a dick, but also I understand the abundance theory to be "it will work to lower prices and then voters will be happy"
23.09.2025 18:28 β π 23 π 2 π¬ 2 π 1Good policy and good politics are often not the same which is awkward
23.09.2025 18:27 β π 13 π 2 π¬ 0 π 0But I also read this as "you should probably not try to explain your specific plan to fix stuff and hope voters like the process" because they don't care (ie, abundance)
23.09.2025 18:26 β π 14 π 1 π¬ 2 π 0Imo it's both salience and positioning. Voters always want things to be magically fixed without tradeoffs, and the way they get information about issues is super messy. However, that info still does come from things that happen in the world
23.09.2025 18:25 β π 14 π 1 π¬ 1 π 0Mm, I think voters do have policy beliefs, even if they're not thought out or coherent. I.e. I think many voters want a reduction in illegal immigration, even if they don't like the policies that would actually get there
23.09.2025 01:40 β π 9 π 0 π¬ 1 π 1Yeah no worries, I do think we have points of agreement
23.09.2025 01:29 β π 1 π 0 π¬ 0 π 0Wait your argument here is Democrats have never actually tried to convince voters to trust them on the economy or immigration? That doesn't scan with the vast quantity of ads and persuasion material I've seen run
23.09.2025 01:16 β π 4 π 0 π¬ 2 π 1Okay, but *they do*. Voters do trust Republicans on these issues (not including climate). You can't just say it's not rational and dismiss it
23.09.2025 01:10 β π 1 π 0 π¬ 1 π 0The existence of negative media isn't a reason to handwave away real disagreements
23.09.2025 01:07 β π 3 π 0 π¬ 1 π 0I don't think this is right. Yes, Republicans constantly focus on the worst Democratic issues, but that happens the other way around as well, and it works best when Dems are out of step with voters. If your position is in step with voters, the other side has less to work with
23.09.2025 01:06 β π 1 π 0 π¬ 2 π 0